Download Financial Transaction Tax

EUROPEAN COMMISSION
Brussels, 28.9.2011
SEC(2011) 1102 final
Vol. 8
COMMISSION STAFF WORKING PAPER
IMPACT ASSESSMENT
Accompanying the document
Proposal for a Council Directive
on a common system of financial transaction tax and amending Directive 2008/7/EC
{COM(2011) 594 final}
{SEC(2011) 1103 final}
EN
EN
ANNEX 7
POLICY OPTIONS
1.
POLICY OPTION 1: FINANCIAL TRANSACTION TAX
Box (1). Key terms and concepts
Algorithmic trading: Automated transactions where a computer algorithm decides the order-submission strategy.
See also: High-frequency trading. (King and Rime (2010))
Capital Duty: A levy based on the issuance of capital, e.g. the emission of new shares or bonds in the primary
market.
Currency Transaction Tax (CTT): A tax that is levied on the gross transaction volume of currency transactions in
spot markets as well as in future and derivatives markets involving currency transactions.
High-frequency trading (HFT): An algorithmic trading strategy that profits from incremental price movements
with frequent, small trades executed in milliseconds for investment horizons of typically less than one day. See
also: Algorithmic trading. (King and Rime (2010))
Financial transaction: In its broadest sense, a financial transaction can be defined as any (contractual) transfer of
financial assets or as any payment of money (cash or transfer) between a buyer and a seller attached to the
(contractual) transfer of a right with an economic value. In principle the transferred right can be a non-financial
good or service as well as financial assets and instruments like equities, bonds, derivative contracts, structured
products, repos or currencies. For the purpose of transaction taxes, financial transactions refer in the context of
this document to transfers of financial assets and instruments only.
Financial Transaction Tax (FTT): In this document a financial transaction tax is defined as a tax which consists
of two elements, namely a CTT and an STT.
Market liquidity: A characteristic of the market where transactions have a limited impact on prices (“price
impact”) and can be completed quickly (“immediacy”). (King and Rime (2010))
Securities Transaction Tax (STT): A tax that is levied on the gross transaction volume of primary and secondary
market transactions with equity securities, debt securities and alike products (e.g. certificates, warrants, units in
funds, structured notes etc.) and related (including commodities) derivative products including options, swaps,
futures and forwards traded in exchanges and over-the-counter (OTC). STT can also be levied on a subset of
these financial instruments.
1.1.
Scope of application: Taxable event and place of taxation
This section discusses the basic concepts and prerequisites for an implementation of a
financial transaction tax as a general framework to assess the impacts of this policy option.
The general concepts outlined below are not related to the precise scope of the tax (CTT, STT
EN
2
EN
or FTT) but touch upon general implementation requirements. Some of the aspects raised here
are also discussed in IMF (2010b).
1.1.1.
Primary vs. secondary market financial transactions
The first question that arises when designing a tax on financial transactions is whether this tax
should be imposed on both primary and secondary market (financial) transactions or only on
one of them. Primary market transactions relate to the issuance of the financial instruments,
whereas secondary market transactions refer to the subsequent transfers of those instruments.
One possibility would be to levy the FTT only on secondary markets. The argument for this
approach is that a transactions tax is targeting trading activity, but not the process of capitalraising by companies and governments. This is also the basic principle of the Capital Duty
Directive (Directive 2008/7/EC). The tax neutrality of restructuring operations is a principle
embedded in both the Capital Duty Directive (for indirect taxation/capital duty) and, for direct
taxation, the EU Merger Directive (Directive 90/434/EEC). It is derived from the need to
remove obstacles to the freedom of establishment and to encourage company cross-border
mobility. Also, restructuring operations referred to in the Capital Duty Directive cannot be
subject to any form of indirect tax according to the current legislation.
On the other hand, it is obvious and there is theoretical and empirical evidence (see also
Annex 12) that the tax burden of a transaction tax will ceteris paribus depend on the trading
frequency of the issued instrument. If only secondary market transactions are taxed,
companies' bonds and equities and governments bonds will depreciate (i.e. the cost of capital
will increase) depending on how often they are traded. If the tax was only applied to the
primary market, this unequal treatment of instruments would be reduced since every
instrument would face the same cost at issuance, but its tax burden would not depend on the
trading frequency afterwards.
Some derivatives and especially almost all Over-the-Counter (OTC) derivatives are not traded
at all.1 They are bilateral contracts between two parties with no secondary market as for
example also loan agreements. If a tax is levied only on secondary markets, these instruments
would remain largely untaxed. If there was the objective to tax transactions in these products
a tax on primary markets – to be more precise on the contracting of the derivative – would
need to be considered. To the degree that they are not used to raise capital, their taxation
would however be possible under the Capital Duty Directive.
Thus, from an economic point of view, in order to avoid further distortions, the alternative to
tax both, primary (contracting) as well as secondary market financial transactions might have
less disadvantages than taxing only primary or secondary market transactions.
1
Note that trading in bond markets has indeed decreased since it has been partly substituted by the use
derivatives. If a market agent wants to transform interest payments received from a bond or hedge the default
risk from a debt security he does not necessarily sell the bonds and purchase assets with the desired interest flow
or risk profile but rather keeps the bonds and buys for example interest rate swaps or credit default obligations.
EN
3
EN
Since the taxation of capital issuance infringes the Council Directive 2008/7/EC of
12 February 2008 concerning indirect taxes on the raising of capital (Capital Duty Directive)
a first legal step necessary for the implementation would therefore be the change of the
Directive. Additional changes would need to be introduced with regard to restructuring
activities of companies. The Directive states that restructuring operations cannot be subject to
any form of indirect tax whatsoever. In order to solve this, an exemption for restructuring
transactions could be foreseen.
1.1.2.
Taxable event
For all definitions below, the basic concept of the taxable event of a transactions tax is the
transfer of ownership of a financial instrument or the registration of a contract in case of OTC
derivatives. Hence, the taxable event is defined as the exchange of the instrument and the
legal transfer of the property right with all connected rights and duties connected to the
ownership. In cases where no exchange of instruments takes place, the taxable event is when
the contract is concluded while the chargeable event is at the point where the legal obligations
occur. For derivatives the taxable event is accordingly the moment when the contract is
agreed upon.
1.1.3.
Place of taxation and taxpayer
In order to define the financial transactions that each jurisdiction would be entitled to tax
under the FTT, different principles could be used. More particularly, the right to tax could be
defined by reference to (a) the tax residence of the parties involved in the financial
transaction; (b) the place where the financial transaction is deemed to have taken place; or (c)
the place of issuance of the financial instrument being traded. In theory, a tax could
potentially employ more than one of the connecting factors at the same time outlined above
(thereby extending the potential tax base), as long as double taxation is avoided. A more
detailed explanation of each of these principles is provided below.
Within these three principles, the taxpayer could be further distinguished to allow for an
application of the tax on certain traders only.
EN
4
EN
Box (2). Proprietary trading
The tax could potentially be limited to two factors – (i) transactions involving proprietary trading by (ii) financial
institutions. Proprietary trading (mainly by investment banks and hedge funds) has been subject to considerable
criticism and is assumed to pose significant risks for single institutions, specific markets and consequently for the
financial system as a whole. Those concerns are also behind the reasoning for the "Volcker Rule" proposed in the
United States that aims to limit the trading that banks do with their own money.
The differentiation between proprietary and non-proprietary trading is however only possible if the regulatory
environment is providing the means to clearly identify these transactions and to share the relevant information
with tax authorities. The information on proprietary trading would be most accessible at the level of the financial
institution, which would normally keep separate track (even separate desks) of proprietary and client trading. The
drive of financial regulation towards increased transparency of financial markets could allow that information to
be at the disposal of financial markets as well. In that respect, any definitions of proprietary trading for tax and
regulatory purposes may need to be broadly aligned.
Proprietary trading in its broad definition can take the form of directional calls (i.e. betting on the market trend),
price/risk arbitrage, market-making, hedging, etc. The risk associated and the market impact of those activities
may differ significantly.
(a) The FTT based on the Tax Residence Principle
(i) Member States' taxing rights
Under a FTT based on the tax residence of the parties involved in the financial transaction,
each jurisdiction would be entitled to impose FTT on every financial transaction conducted by
individuals and entities that are considered as residents in this jurisdiction, regardless of
where the financial transaction is deemed to have taken place. Under this principle, the market
participant would therefore be liable to tax for all his or her global financial transactions in
her jurisdiction of residence. The correct determination of the tax residence of market
participants must in this case be clearly defined in order to avoid double or non-taxation. This
points to the well-known problem of determining the place of tax residence of taxpayers in an
international context.
In order to apply this principle it would be necessary to clearly identify the legally relevant
buyer and seller of every financial transaction – the persons or companies which actually
exchange the property rights and receive the economic benefit from the trade. The tax would
be levied on one or both parties involved in the financial transaction, depending on whether
they are resident for tax purposes in a jurisdiction applying the tax. If one leg of the trade is a
non-resident buyer/ seller, he would accordingly not be taxed. If both parties are resident for
tax purposes in jurisdictions levying the FTT, the FTT would be a two-legged tax levied at
half the total rate for each leg.
The FTT based on the Tax Resident Principle would thus allow the state of tax residence to
tax all the financial transactions entered into by its tax residents, regardless of where the trade
takes place. By contrast, tax residents from countries without such FTT would not be taxed.
EN
5
EN
A strict application of this system could mean that investors from countries that do not levy
the tax would be tax free and this would create a tax wedge: Non-EU investors might have a
higher incentive to invest in a European company/ government bond than domestic residents.
In addition, exchange of information with third countries might be necessary2.
(ii) Tax collection issues
While this principle has advantages, notably that the tax revenue would be distributed
according to the financial trading conducted by tax residents, the enforcement and compliance
costs must be addressed for mainly three reasons:
a. EU intermediaries
Firstly, the cost of collecting the FTT might be very high. The collection would be possible
only if, either every tax resident (private households, companies, government agencies)
reports to the tax authorities the financial transactions in which they enter; or, administratively
easier and less costly, if financial intermediaries like banks, brokers, funds, exchanges etc. are
considered as the collecting agents of the tax.
If the financial intermediary involved in financial transactions effected by residents in
jurisdictions where the tax applies is from an EU Member State, it should be in charge of
collecting the FTT. There are, however, practical issues with applying that approach. In the
first place, intermediaries may not necessarily be in a position to determine correctly the tax
residence of their client and to assign the tax revenues to the appropriate Member State. In
addition, for some derivatives, as outlined in 5.2.2.2., there may be no consideration (i.e.
payment) at the time of contracting, and therefore, no monetary flow from which the tax could
be withheld/ collected.
A tax only on proprietary trading of financial institutions would arguably not be as
challenging from tax collection point of view, because financial institutions would keep better
track of their investments and their economic substance than retail investors.
b. Non-EU intermediaries
If an intermediary in a non-taxing jurisdiction is used, the tax resident is liable to tax. For
OTC markets, the tax would be collected when contracts are registered with an official
regulatory body (central clearing platforms, trade repositories or other bodies) and only if one
of the contracting parties is a domestic tax resident. This would require a high degree of
international cooperation on exchange of information with countries where the intermediaries.
In addition, the administrative costs for the tax authorities for tracking transactions carried on
a global basis may be excessive.
2
EN
Such systems do already exist. An example is the EU Savings Directive (2003/48/EC).
6
EN
c. Non-EU financial markets
In order to impose the FTT on transactions conducted by residents in jurisdictions where no
tax is applied, the national tax administration may need to receive the information on the
details of the transaction taking place abroad not only by non-resident intermediaries, but also
by the non-resident financial markets too. This would also require not only a high degree of
international cooperation on exchange of information, but also a degree of transparency (and
development of that degree of transparency over time) of those non-EU OTC markets that is
equivalent to the one in the EU.
d. Relocation
In terms of relocation it should be noted that there is the possibility that a company moves its
trading activities to subsidiaries or other related entities in a third non-taxing country and
repatriates profits from trading via dividends (which would be exempt either under the nontaxing country's domestic law or under its double taxation agreements), or via transfer pricing
arrangements.
Financial transactions carried out on exchanges or where central counterparties are interposed
in order to reduce counterparty risk and increase transparency often involve a number of
parties and intermediaries. One transaction may be broken down into a series of transactions
first between the initiating party and the broker/ dealer who further deals with a member of a
clearing house (exchange) and who in turn deals with the central counterparty (CCP)/ clearing
house (exchange). The other side of the transaction may be divided in the same number of
transactions. In this case for each transaction in the market would need a (cross-border) lookthrough approach in order to identify the legally relevant agents. A solution to that may be the
assignment of withholding/collecting and reporting duties to the intermediary that is
immediately acting for the originator of the transaction.
The Anti-Money laundering regulations could be helpful tools to deal with this problem. The
advantage of the look-through approach is that buyer and seller could be identified and their
nationality or the country of residence for taxation could be disclosed. This would be
necessary if revenues are to be shared among the governments involved.
(b) The FTT based on the Source principle
Under a FTT based on the Source Principle, each Member State would have the right to tax
all the financial transactions that are deemed to have taken place in its jurisdiction, regardless
of the tax residence of the parties involved in the transaction. The Swedish transaction tax
experience can be seen as a tax levied according to this principle. Alternatively, the place of
settlement could be defined as the tax location. Note that the place of settlement and the place
of transaction do not necessarily coincide. This might have effects on enforcement and
revenue distribution.
Designing the FTT in this way would require defining criteria to determine where a financial
transaction is deemed to have taken place. For instance, OTC contracts could be taxed in the
jurisdiction where the derivative/ trade is registered or settled, regardless of the tax residence
EN
7
EN
of the parties to the contract. Financial transactions using exchanges or any organised trade
platform can be considered to take place in the jurisdiction where these centralised systems
are located.
Carefully defining on whom the tax would fall legally and when the tax can be collected is
necessary. For instance, it must be defined whether the FTT is to be paid at the moment of
notification of the trade or at the moment of settlement of the contract. Equally, the exchange,
settlement system, or central information points which would be legally bound to collect the
tax must know which person is liable to pay the tax, the buyer, the seller or both. A possibility
would be that these central collecting agents debit the buyer and the seller of each transaction
with 50% of the tax.
The advantage of designing the FTT under this principle is that no identification of the party
instructing the transaction is necessary. The tax could be collected at central clearing
platforms or electronic exchanges and other trading platforms thereby avoiding excessive
administrative costs.
Nevertheless, there are some other design issues to be considered, namely:

It would need to be assured that transactions within a business entity are also taxed,
to exclude tax avoidance by vertical as well as horizontal integration (conducting
transactions within rather than between businesses): The result could be larger
financial institutions creating possibly additional systemic risks.

The fact that a central tax collection system exists would lead to a geographical
concentration of the revenue in countries with major trading centres. Given the
strong concentration of financial markets in some Member States the distribution of
revenue might be very uneven. Systems of revenue-sharing among countries could
also be agreed, but not without related data collection for the sharing.

The risk that the transactions together with the trading platforms are relocated to
jurisdictions which do not levy such a tax is very high.
(c) The FTT based on the Domestic Issuance Principle
Under a FTT based on the Domestic Issuance Principle, each Member State would have the
right to tax the financial transactions involving securities issued by domestic actors, notably
shares and bonds by domestic companies, domestic government bonds and derivative
contracts settled or contracted in the country.
This principle is similar to the basic idea of the UK stamp duty where enforcement is
strengthened by linking evidence of ownership to the proper payment of the duty. The UK
stamp duty taxes only shares of UK-registered firms irrelevant where the trade takes place
(see also Annex 8). In the UK the regulations permit a paperless transfer of shares to be
registered, provided it is made through an electronic system approved by the Treasury under
the regulations. The SDRT regulations impose an obligation on the operator of CREST (or
EN
8
EN
any other Treasury approved electronic transfer system) to collect SDRT on transfers going
through its system The responsibility for paying SDRT rests with the “accountable person” as
defined in the SDRT Regulations. This is usually the broker acting for the purchaser of the
shares. Where the SDRT is not paid through CREST, the customer is required to send full
details of the transaction to the SDRT Office with payment of the SDRT. Once payment is
received an official receipt for the payment is issued and any interest on late payment
calculated and requested.
1.2.
Taxable base
The heterogeneity and rapid evolution of the different financial instruments pose difficulties
for the definition of the taxable base of the FTT.
A useful tool to define and track the products covered by the FTT would be a harmonized
nomenclature of financial instruments and products in order to guarantee that for tax (and also
regulatory) purposes, Member States use equal definitions. A role model could be the
Standard International Trade Classification used in the area of customs. All existing financial
instruments should be listed in order to make sure that all product definitions are identical in
all countries. Without such a tool there is large scope for interpretation with regard to the tax
treatment of certain products which could in turn lead to substitution and loopholes.
Nevertheless, it must be borne in mind that employing exhaustive lists would arguably
facilitate arbitrage around strict inventory of definitions. Changes to such a list would need to
be agreed by Member States.
1.2.1.
Spot transactions
The definition of the taxable base for spot transactions should not pose serious problems.
There is a large amount of national experience and the value of the asset as priced in the
transaction could be taken as the taxable amount. The gross transaction volume would be the
tax base.
1.2.2.
Derivatives
Defining the FTT base for derivative transactions raises more questions. It has been proposed
that the value of the underlying instrument or asset, i.e. the notional value, is used as the
taxable base. This value is the nominal or face amount that is used to calculate and record
payments on the instrument. This notional value is commonly used in options, futures and
currency markets and for a large share of derivatives easier to observe than e.g. prices or
premiums which for some products do even not exist.
There is also a more fundamental economic argument for the use of the notional as the tax
base. If the FTT were based on the premium, traders are essentially rewarded for using more
out-of-the-money options. These products have disproportionately lower premiums due to the
higher implicit leverage. For example, if the premium is the tax base traders could switch
from an option with the same notional value but a different strike price and reduce the
EN
9
EN
premium and therefore the tax burden. When using the notional the tax will be invariant to the
amount of implicit leverage in the option.
However, the notional value as tax base poses also a number of questions. It is a fictitious
reference and may not be taken as the real value of the derivative contract insofar as this
underlying asset/instrument is very often not transferred at all. In addition, the link between
the value of a derivative and its underlying varies widely among the different types of
derivatives, but also within the same type of derivative depending on its terms. Finally, as will
be discussed in section 6 using the notional could lead to significant market reactions which
have to be taken into account.
Box (3). Derivatives – some background
In financial terms, a derivative is a financial instrument - or more simply, a contract whose performance is based
on the behaviour of an underlying asset. The price of a derivative is thus derived from one or more underlying
assets. It is a financial contract with a value linked to the expected future price movements of the asset it is
linked to - such as a share or a currency. Thus, a derivative is an agreement between two parties that is
contingent on a future outcome of the underlying.
The underlying could be anything but the most common are commodities, securities, currencies, interest rates,
yields, a stock index or other financial measures, transfers of credit risks, climatic variables (i.e. weather), freight
rates, emission allowances, inflation rates and economic statistics. The overall derivatives market has five major
classes of underlying assets: interest rate derivatives (the largest), foreign exchange derivatives, credit
derivatives, equity derivatives and commodity derivatives.
There are many kinds of derivatives, with the most notable being swaps, forwards, futures, and options. Since a
derivative can be placed on any sort of security (or indeed behaviour or event), the potential base of derivative
contracts is very large. Even derivatives themselves can be used as an underlying asset. Moreover, derivatives
can take the form of very complex contracts and are not always based on assets with a direct value to the price of
the derivative. For example, weather derivatives are financial instruments that can be used by organisations or
individuals as part of a risk management strategy to reduce the risk associated with adverse or unexpected
weather conditions. The difference from other derivatives is that the underlying asset (rain/ temperature/ snow)
has no direct value to the price of the weather derivative.
Furthermore, compared to bonds and equity, a distinction between primary and secondary market transactions is
less relevant, but derivative contracts can also be distinguished by the way they are traded in the market, i.e.
over-the-counter (OTC) or exchange (organised markets) traded. OTC or off-exchange trading is to trade
financial instruments such as derivatives usually/ traditionally directly between two parties. It is contrasted with
exchange trading, which occurs via facilities constructed for the purpose of trading.
Despite these difficulties, to the extent that the notional value would in many cases be the
only value which is readily observable in a derivative contract, it would necessarily be the
natural starting point for defining the tax base of the FTT, at least for futures, options and
forwards, as more fully explained below:

EN
Futures/ Forwards: A futures contract is an agreement to buy or sell a financial
product or commodity at a reference price (future price or strike price) at a specified
10
EN
date in the future. Futures are standardized with regard to the quantity and quality
and are traded on exchanges. Forwards are similar products but contracted over-thecounter (OTC). The tax base for these contracts would be the underlying asset
evaluated at the strike price. However, in case these derivative contracts are actually
executed, the question arises whether a new taxable event occurs. If that is the case
the underlying financial asset would be taxed with FTT (VAT in case of
commodities) at that moment at the price stipulated in the contract and double
taxation (first taxation is on the derivative contract as such) would occur. As a result
any determination of the tax rate applicable to the notional value of derivative should
take this into account.

Options: The owner of an option has the choice (but not the obligation) to buy or sell
a financial product or commodity at a reference price (exercise price) at a specified
date in the future; the owner exercises the option if it would be advantageous for him
to do so. The underlying evaluated with the reference price (notional value) would be
used as the taxable base. For example, one S&P 500 Index futures contract obligates
the buyer to 250 units of the S&P 500 Index; if the index is trading at $1,000 (given
price), then the single futures contract is similar to investing $250,000 (250 x
$1,000). Therefore, $250,000 is the notional value underlying the futures contract.
However, in case the option is exercised the question arises whether a new taxable
event occurs. If that is the case, the FTT tax base relating to the transaction of
acquiring the underlying financial asset would be the exercise price (normally VAT
taxation in case of commodities) and it would be logical to only apply FTT to the
premium paid in respect of the option contract as such.

Swaps: Defining the notional value for swaps might be more difficult, since easy
transformations can decrease their value. For example, interest rate swaps are
agreements to exchange, over a period in the future, a series of fixed interest rate
payments for a series of variable interest rate payments. The notional value of an
interest rate swap could be divided by an arbitrarily large number while the interest
payments are multiplied with the same number. The contract would be economically
the same but the tax base defined as the notional would be close to zero. In this case
a substance over form or some other legislative limits have to be introduced in order
to avoid these effects. In the case of a credit default swap (CDS) - which is a swap
contract in which the purchaser of protection (the CDS) makes a series of payments
to the protection seller and, in exchange, receives a payoff if a credit instrument
(typically a bond or loan) defaults the base could be the insured underlying.
The examples above illustrate that the notional value can be used as the taxable base of the
FTT in certain financial transactions. However, the notional value is difficult to determine in
some other transactions (for instance the weather derivatives mentioned in box (5) credit
default swaps or structured products if derivatives are involved). In these cases, other proxies
for the tax bases have to be found. In these cases premiums paid for the contract or on the
value to be paid in case of the occurrence of the “insured” event could be used.
In any case, the multitude of financial products would need to be filed and catalogued in order
to have an internationally agreed tax base for each product category. If the notional is not
applicable alternative other values could be considered which can serve as a taxable base.
EN
11
EN
Box (4). Issues specific to structured products
Structured products are similar to derivatives in the sense that they are market-linked products. There are
multitudes of product categories within the category of structured products, allowing different “packaged”
investment strategies. For example, a structured security may take the form of a note that provides varying
degrees of capital protection (e.g. from 30% to 100%, but also above). Such types of products have become
increasingly popular with retail investors. The investment is essentially split into two parts. One part (e.g. 90 of
an investment of 100) is invested in a zero-coupon bond that would yield 100 upon maturity of the structured
note, while the remaining 10 are used to purchase derivatives (e.g. options and/or swaps) to track a reference
security/basket/market. The leverage of those derivatives allows significant (contingent) returns on the note,
combined with capital protection.
A simple approach in taxing structured products of the type outlined above would be to tax the investment of
100. Nevertheless, such an approach would omit the nature of the embedded derivatives and the potential
leveraged profits associated. It would also pose an issue of lack of consistent treatment of self-standing and
embedded derivatives.
Normally, accounting rules would have to “pierce the veil” of the structuring by applying a “bifurcation”
approach, reporting the embedded derivatives separately. Such an approach may be applicable for corporate
investors that apply accounting rules, but does not appear practical with regard to individual retail investors.
The above example is provided for illustration purposes only. In other cases the embedded derivatives may not
be that easily identifiable or measurable.
1.3.
Exemptions
Financial intermediaries: clearing members, clearing house, brokers could be exempted for
reasons of tax neutrality and to preserve liquidity in the market. This would however reduce
taxable volumes significantly.
In addition, hedging activities through derivatives may be seen as disadvantaged from the
levying of a FTT based on the underlying. Although there may be arguments that a very low
tax rate would not have a significant impact, the actual tax paid may be substantial compared
to the ‘price’ paid for the hedging (e.g. the premium in case of options). Therefore, an
exemption may need to be implemented, e.g. through a refund for hedging activities. The use
of hedge accounting by the contracting party with regard to the specific derivative may be an
indication which could potentially be used for the exemption if the FTT is based on the Tax
Residence Principle.
A tax levied only on proprietary trading would effectively exempt all non-financial
institutions and private investors.
EN
12
EN
1.4.
Types of FTT
The tax principles and tax bases outlined above can be applied to different product scopes. In
the recent discussions three transactions taxes have been in the focus and will be the variants
of this policy option discussed in this Impact Assessment.
1.4.1.
Option 1A: Currency Transaction Tax (CTT)
A tax on currency transactions is based on the initial idea for a transaction tax on foreign
exchange markets by Tobin (1974, 1978). He argued that the increased mobility of private
financial capital - especially after the end of the Bretton Woods system - might lead to
excessive shifts of funds that create real economic costs for national governments and
economies. Tobin reasoned that the tax could increase the effectiveness of domestic monetary
policy. This document uses the name CTT and not the expression ‘Tobin Tax’ which is
sometimes used in the popular debate.3
The basic idea of the CTT is to levy a tax on currency transactions and related instruments
(FX forwards and FX swaps) transactions. The collection would take place centrally at
currency exchange systems, namely in Real Time Gross Settlement (RTGS). A real-time
gross settlement system (RTGS) is a payment system in which processing and settlement take
place continuously (‘in real time’) rather than in batch processing mode. Like this,
transactions can be settled with immediate finality. ‘Gross settlement’ means that each
transfer is settled individually rather than on a net basis. In the EU such a system is Target2
(Trans-European Automated Real-time Gross Settlement Express Transfer System), which is
operated by the Eurosystem and comprising 26 Member States. For transactions which are
tracked in such systems a tax could be levied centrally.
Proposals for such systems have been made by the Report of the Committee of Experts to the
Taskforce on International Financial Transactions and Development for the Leading Group of
countries (2010), Jetin and Denys (2005) and Spahn (2002). A recent study on revenue
estimates based has been published by Schmidt (2008). Compared with other transaction taxes
the technical discussions are more advanced for the CTT. In this respect, it should be noted
that many proponents and notably the report of the Leading group of countries explicitly
favour a global application of the CTT.
Also, note that this option raises legal concerns which would need to be addressed by
adapting core elements of the TFEU. See Box (5).
3
Tobin himself argued a few months before his death that his name was misused in the debate on financial
transaction taxes. See http://www.econ.yale.edu/news/tobin/jt_01-09-02_ds_misusing-name.htm for an interview
with Tobin on this.
EN
13
EN
Box (5). Legal issues with a CTT
A Currency Transaction Tax, i.e. a levy the taxable event for which is the exchange of currencies, indirectly
restricts the underlying transactions, both between Member States of different currencies and between Member
States and third countries, by rendering them more costly. This may affect payments for the supply of goods or
services (current payments) and investments made, for example, by pension funds (capital movements).
Although the levy would not apply to the cross-border flow of money as such, it would, absent similar effects on
purely national flows, restrict free movement of capital, within the meaning of Article 63 TFEU. This provision
commits not only Member States, but also the Union.
Nothing can justify this restriction, since the cross-border flows affected are not objectively different from purely
national flows (or flows within a single currency zone [i.e. the euro area]), nor could any overriding reason
relating to the public interest serve as a justification. Even if e.g. raising funds to benefit stability funding were to
be considered as an overriding requirement of general interest, that requirement could not explain why
transactions involving countries with different currencies would be treated less favourably than those involving
only one currency. Furthermore, the tax is considered to be disproportionate as funds could alternatively be
raised by other means of budget attribution without affecting a basic freedom of the Treaty and, in any event,
because the scope of the tax would be unrelated to the risks to be covered by the tax revenue raised. Even a very
low tax rate would constitute an infringement, and it would not be possible to establish a threshold of
insignificance.4
As regards Council Directive 2008/7/EC, its Article 5 (2) provides that Member States shall not be subject to any
form of indirect tax the creation, issue, admission to quotation on a stock exchange, trading with stocks, shares
or other securities of the same type, or of the certificates representing such securities. This concerns also loans,
including government bonds, raised through the issuance of debentures or other negotiable securities, or any
formalities relating thereto. Article 6 (1)(a) of the Directive 2008/7/EC expressly states that “[n]otwithstanding
Article 5, Member States may charge duties on the transfer of securities, whether charged at a flat rate or not.”
It would need to be assessed whether these rules of Directive 2008/7/EC would oppose the introduction of
currency transaction levy, in which case its introduction would require a corresponding change of the Directive.
1.4.2.
Option 1B: STT without derivatives and currency transactions
The STT without derivative and currency transactions corresponds to the narrow-based
transaction tax described in the SWD on financial sector taxation which would tax only the
spot transactions of equities and bonds. It would be levied on all bonds and stock transactions
executed in regulated markets thereby making the technical implementation cheap and easy
since these systems are operated with centralized economic systems. This option is similar to
the UK stamp duty when combined with the domestic issuance principle. Note that for the
bond trading product substitution must be considered. It is easy to replace fixed-interest
securities with standard bank loans or deposits which are exempt according to the current
proposals.
4
On free movement of capital and CTL, cf. also Opinion of the ECB on the 4 November 2004 at the request of
the Belgian Ministry of Finance on a draft law introducing a tax on exchange operations involving foreign
exchange, banknotes and currency (CON/2004/34). http://www.ecb.int/ecb/legal/pdf/en_con_2004_34_f_sign.pd
EN
14
EN
1.4.3.
Option 1C: FTT
This option would cover all financial transactions as outlined above. In fact, it consists of the
STT and the CTT. Since it would cover also OTC markets a full coverage of transactions and
access to all transactions data is necessary to effectively be in a position to levy the tax
independently of the tax principle chosen. Figure (1) shows the products covered in Option
1C.
Figure (1): Variants of an FTT
FTT
Gross transaction volume/ Derivatives: Notional value
Shares
Bonds
Futures
Derivatives
(OTC and exchange)
Currency Transactions
1.5.
The tax rate
Generally, the level of the rate would depend on factors such as the revenue potential, the
level of impact on targeted markets or the importance of the tax rate relative to the transaction
costs in specific markets. There is obviously a link between the tax base and the tax rate of the
FTT. Thus, if, as illustrated above, the notional value of a derivative contract is taken as the
tax base but this value does not reflect the real value of the product at stake, in order to avoid
EN
15
EN
excessive taxation of such a product, a “lower” tax rate (or a range of “lower” tax rates) is
advisable.
With regard to the structure of the tax rate, two options are possible: a flat rate and
differentiated rates.
1.5.1.
Flat rate
One option is to set a uniform rate for all transactions. This has the merit of being simple and
easy to comply with, but it might not sufficiently take into account the specifics of the
products traded and the potential impacts on the different markets.
Most recent proposals assume a low statutory tax rate which ranges between 0.01% and 0.1%.
An exception is the above mentioned proposal for a CTT by the Committee of Experts to the
Taskforce on International Financial Transactions and Development for the Leading Group of
countries (2010). They explicitly want to avoid market distortions and suggest a tax rate as
low as 0.005%.
By contrast, some Member States have experience in applying stamp duties on equities
amounting to 0.5%, and even 1.5% to address tax avoidance. Note that these duties do in
general not include derivatives in its scope. However, a tax rate of 0.5% might have much
stronger effects if it is applied to notional values in derivatives markets than in equity
markets. Thus, a relatively low level of taxation is likely to be essential to avoid strong
negative impacts on markets and to ensure some revenue collections, since the incentives for
avoidance increase with the tax rate. (See also Annex 10 for a detailed description of the
effective tax rates and other economic issues.)
1.5.2.
Differentiated rate
A second option is to set differentiated rates, depending on several factors.
EN

Varying tax rates could be based on the impact of different tax burdens of financial
instruments on markets and the (political or economic) desirability of the instrument.

Tax rates could also be differentiated according to the type of counterparty: banks,
other financial institutions (such as hedge funds) for their proprietary trading and
non-financial corporations, for example if one makes the assumption that certain of
these categories are more prone to speculative trading than others or if reductions in
trade volume are different. This information would have to be gathered for all traders
in financial markets.

Another approach might be to look at existing transaction costs and spreads in the
various financial market segments. Indeed, these transaction costs will determine the
impacts on market liquidity and a transaction tax would clearly be a (new) part of
these costs. In markets with high transaction costs (less liquid assets) a single rate is
likely to impose a much smaller percentage increase in the transaction costs than
where these are low (more liquid assets). In the first case the transaction tax will
16
EN
have less impact on liquidity. Hence, tax rates might be defined relative to existing
transaction costs to avoid large impacts on markets. This would of course imply the
exercise of defining the current level of transaction costs for different markets
(products) or market segments, both exchange (organised markets) and OTC. An
update mechanism relying on unanimity would need to be set.

Pollin et al. (2002) proposed the following scale of tax rates for a possible Security
Transaction Tax (STT) for the US based on the principles of taxing the value of the
asset being traded which does not necessarily equal the notional value, and of taking
into account the existing transactions costs by adapting the tax rate to the transaction
costs level. The proposal takes the 0.5% rate of the UK stamp duty on equities as the
benchmark for establishing rates in other markets in a way that minimises
distortions:
–
Bonds: 0.01% per each year until bond's maturity.
–
Futures: 0.02% of the notional value of the underlying asset.
–
Options: 0.5% of the premium paid for the option.
–
Interest Rate Swaps: 0.02% per each year until maturity of the swap
agreement.5
–
Foreign exchange: 0.01% on each spot transaction.
Box (6). Technical aspects dealing with relocation
Although not directly linked to the technical feasibility of a transaction tax, it will be of vital importance for the
potential success of any FTT to avoid (or to assess the risk of) relocation of transactions to havens where the tax
does not apply - if the tax is not introduced in all countries where major (or to become major) financial centres
are located.
First, if Member States decide to tax according to the Residence Principle or the Domestic Issuance Principle,
relocation for tax purposes might be pointless in the theoretical case of full and cheap enforcement. In reality,
these taxation principles will however raise problems of implementation and of enforceability in the absence of
well functioning administrative cooperation and strict reporting obligations to national authorities by domestic
intermediaries (source principle) as well as domestic and foreign intermediaries/ tax administrations (residence
principle). In case of a unilateral introduction of the tax at EU-level information sharing mechanisms with third
countries should be implemented if the residence principle is envisaged.
Second, there are already differences in transaction costs and spreads between countries and the impacts of these
and the increase of existing transactions costs due to the introduction of a transaction tax seem to be a
determinative factor to assess possible relocations. How far can transaction costs be stretched before relocation
occurs? Successful (derivatives) markets enjoy a combination of characteristics and low transactions costs are
only one of them. Other factors include ready access to capital, a solid regulatory regime, legal certainty and
5
Baker, Pollin, Mc Arthur and Sherman (2009) took 0.01 percent for each year until maturity for the sake of
revenue calculations following the pattern with bonds.
EN
17
EN
political stability, and a large pool of professional talent.
Third, some market places might have a unique position or benefit from “network externalities” which would
outweigh the burden of the tax so that it is still an attractive venue for trade. Some researchers refer to the
extreme concentration of transactions on exchanges in Europe (only 6% are spot transactions, 94% refer to
futures and options) to show that network externalities are well established. However, it can be expected that
rivalry amongst the world’s leading financial centres is very intense and market concentration might change
quickly. This is particularly important for an application of the tax with limited geographical coverage. If it is
considered to tax at EU-level only, it seems indicated to include at least the major financial centres in Europe,
i.e. London, Zürich, Frankfurt, Geneva, thus including Switzerland.
1.6.
Double taxation issues
As outlined above, States could decide to tax according to the (a) Residence Principle (b)
Source and Territoriality Principle (c) Domestic Issuance Principle. In an extreme case all or a
mix of these principles may be applied at the same time by the taxing States. In this case the
potential for double taxation is obvious if the tax is applied in more than one state. More
specifically, if for example a Member State A applies the source principle while another
Member State B the domestic issuance principle, double taxation might occur when securities
with origin in Member State B are traded in A. On the other hand, if A applied the domestic
issuance Principle whereas Member State B would apply the tax resident principle, nontaxation of Member State's B equity traded by tax residents of Member State A would occur.
Therefore it is again necessary that governments which consider the introduction of such taxes
co-ordinate in order to avoid double or non-taxation.
The different principles needed to establish the taxation competence of States could be
combined in a way that double or non-taxation is minimized to the largest extent possible and
that revenue is more equally spread between countries.
2.
POLICY OPTION 2: FINANCIAL ACTIVITIES TAX (FAT)
2.1.
Rationale
The FAT can best be defined as a “group of taxes”; these taxes share a certain common root
but may nevertheless differ considerably. In essence, the FAT would be levied on the sum of
profit and remuneration of financial institutions, since one common rationale behind all the
forms of FAT is the fact that would a rent occur, it will go either to shareholder in terms of
higher profit (dividends or capital gains) or to workers (via higher remuneration). However, as
more fully explained below, a FAT can take several possible forms depending on how profit
and remunerations are defined and which objectives are pursued with the introduction of the
tax.
2.2.
The FAT as a direct or indirect tax
Whether the FAT is interpreted as direct or indirect tax has important effects with regard to its
potential legal base as well as its integration with other taxes. Broadly speaking, a FAT would
EN
18
EN
be imposed on value added. This is, similarly to a tax on value added; the FAT would aim at
taxing the difference between (i) the proceeds of providing financial services and (ii)
purchases of non-labour inputs of financial institutions. This would be equal to the sum of
profits and labour costs of said financial institutions. However, the fact that the FAT and the
value added tax share the same underlying economic principle and that it is argued that the
FAT could serve to address the current VAT exemption of the financial sector, does not
necessarily mean that the FAT is an indirect tax.
The classification of the FAT as a direct or indirect tax does not merely belong to an academic
debate. As more fully explained below, the definition of the scope of the tax, the methods to
avoid double taxation, the interaction between the FAT and other taxes, and even the choice
of the legal basis of the TFEU which would eventually allow for a harmonised introduction of
the FAT in the EU, depend on whether the FAT is considered a direct or indirect tax. In this
respect, there is not a unique definition of what a direct or indirect tax should be, but, rather
several criteria which are often used to classify a certain tax as a direct or indirect, as follows:

whether the taxpayer is or is not the person on whom the economic burden of the tax
is expected to fall;

whether the assessment of the tax takes into account the circumstances of individual
taxpayers;

whether it is a ‘personal’ tax or ‘in rem’ tax;

whether the tax is levied at regular intervals on sources of income such as
employment or property (direct taxes), or on producers in respect of the production,
sale, etc. of goods and services, which they charge to the expenses of production
(indirect taxes);

whether the tax is a tax on income (including capital gains and net worth) (direct) or
on consumption (indirect).
Under any of these criteria above, the FAT seems closer to a direct tax than to an indirect tax,
since (i) there would not be a legal obligation to pass the FAT on to the consumer of financial
services on a transaction-per-transaction basis; (ii) the assessment of the FAT would take into
account the circumstances of the financial sector; and (iii) the FAT would be periodically
levied on all entities that can be classified as financial institutions, on the amount of profits
and remuneration that they obtain and pay, respectively, over the relevant period.
Accordingly, even if the FAT could be considered as a ‘hybrid’ tax, sharing features of both
direct and indirect taxes (particularly the addition method FAT), for the purposes outlined
above, it will be considered as a direct tax.
EN
19
EN
Scope of application – the geographical dimension of the FAT
2.3.
Proposals have been made to design the FAT, particularly the addition method FAT or FAT1,
in a way that would essentially mirror the origin/destination principles ruling value added
taxation.
Box (7). Design of a tax on the basis of the origin/destination principles ruling value added taxation
1.
An origin-based tax
A first option is to design an origin-based tax. This tax would be levied on the profit from domestic production.
The base would therefore be the proceeds from domestic and foreign sales minus the cash disbursements on
purchasing domestic and imported inputs (including capital goods). An origin-based tax would normally allow
for the deduction of labour costs, however, insofar as the FAT would also be imposed on (excessive)
remuneration for labour, (at least part of) the labour costs would not be deductible.
2.
A full destination-based tax
A second option is to design a full destination-based tax. This tax would be levied only on the profit from
domestic sales (i.e. on the profit of sales to domestic consumers both by domestic or foreign companies). The tax
base would therefore be the proceeds from domestic sales minus the costs on the inputs (material and capital)
used for the production of the goods and services sold domestically. A method of apportionment would
determine the overhead costs to be considered in each country. A destination-based tax would normally allow for
the deduction of the labour costs that are attributable to the domestic sales, however, insofar as the FAT would
also be imposed on (excessive) remuneration for labour, (at least part of) the labour costs would not be
deductible.
3.
A VAT-type tax
A third option is to design a VAT-type tax. This tax would be levied on the profit from domestic sales (i.e.
considering the proceeds of sales to domestic consumers both by domestic or foreign companies). The tax base
would be the proceeds from domestic sales minus the costs of domestic purchases. Similar to VAT, export sales
would not be taxed but all imports would be. A destination-based tax would normally allow for the deduction of
labour costs, however, insofar as the FAT would also be imposed on (excessive) remuneration for labour, (at
least part of) the labour costs would not be deductible.
A destination-based FAT
It is argued that a FAT matching the destination basis of value added taxes would best address
the VAT exemption of financial services. Under a destination-based FAT, neither borrowing
from nor lending to non-residents would be taken into account for the purposes of assessing
the FAT payable by a given financial institution.
However, implementing an FAT in such a manner would require that financial institutions
keep separate accounting track of domestic and foreign financial transactions. Furthermore, it
EN
20
EN
would require defining criteria to distinguish domestic and “foreign” financial transactions
(by reference to the borrower, the place of establishment, etc). To the extent that the foreign
transactions would not be taxed under the destination-based FAT, the risk arises that financial
institutions attempt to maximise their “foreign” transactions in order to avoid paying FAT
(relocation risks). Finally, it would very likely oblige financial institutions to register for FAT
purposes.
An origin-based FAT
Under an origin-based FAT, borrowing from and lending to non-residents would also be taken
into account. This origin-based FAT is also sometimes referred to as the source-based FAT.
Indeed, to the extent that the FAT can be better classified as a direct tax, which is levied on
the institution and not on the transaction, it should be possible to use the concepts of source
and residence to define the scope of application of a sort of origin-based FAT.6
A residence/ source-based FAT
Designing the scope of application of the FAT on the basis of residence and source principles
would have as an advantage that all EU Member States have wide experience on applying
these principles in an international setting. The FAT would only be an additional direct tax. At
the same time, international direct taxation is far from being a fully harmonised area of
taxation, or absent of conflicts. The FAT could be designed to minimise some of these
conflicts but, to the extent it is built up on the principles of international taxation principles, it
would also certainly share some of them.
A FAT designed in accordance with the residence principle would be levied on the worldwide
profits of the financial institutions that are resident for tax purposes in one of the EU Member
States. This would mean that the income and expenses of financial transactions with nonresidents would also be considered for FAT purposes. A method to avoid double taxation
must be envisaged so that the dividends of EU subsidiaries are not subject to FAT in more
than one EU Member State (see below).
In order to better proxy an origin-based FAT, the residence-based FAT on EU domestic
institutions should be coupled with a source-based taxation of EU branches (i.e. permanent
establishments) of non-EU financial institutions. These would be subject to FAT in the EU
Member State where they are located.
An origin-based FAT would probably also require that non-EU financial institutions, whose
activity in the EU is not intense enough to give rise to a permanent establishment, pay FAT on
their financial transactions with EU residents. The best way of applying the FAT in these
cases would be a withholding tax. However, withholding FAT on a transaction-pertransaction basis for non-EU residents could conflict with the Double Tax Conventions (DTC)
6
It is difficult to identify the residence/source principles of direct taxation with the destination/origin principles;
the OECD draws a parallelism between the residence principle and the destination principle; and the source
principle and the origin principle, even if it is acknowledged that residence and consumption are not equivalent
and that source is associated with income, while origin is associated with production. See OECD (2007) Box 7.1.
EN
21
EN
following the OECD Model Convention. If the return to the financial transaction made by the
non-resident financial institution in the EU is classified as a dividend or an interest, the
withholding tax might be allowed under the DTC. By contrast, a withholding tax would likely
contravene most DTC if the income earned by the non-EU resident is classified as business
income (or even if it is classified as a capital gain, depending on the specific DTC).7
2.4.
Taxpayer
The FAT would be levied on corporate taxpayers that can be classified as financial
institutions. The FAT should be levied on a range of financial institutions as wide as possible
in order to avoid that the financing activity shifts to the quasi-financial sector. The financial
sector would certainly include banks, credit card companies, insurance companies, consumer
finance companies, management fund companies, stock brokerages, investment funds, hedge
funds and some government sponsored enterprises, regardless of whether these entities are or
not subject to Corporate Income Tax in their respective Member State of residence.
It has been proposed that all the enterprises conducting more than a certain threshold of
financial activities become subject to the FAT. Such a threshold could be defined, for
instance, in relative turnover terms (turnover of financial activities as compared to the total
turnover of the enterprise, or using the VAT pro-rata of the enterprise as benchmark).8 Such
an approach would allow taxing also intra-group financing and shadow-banking activities.
Referring to the weight of the financial activity as compared to the total business activity
could be the only criterion to define the FAT taxpayer or could be used in combination with a
list of entities that would in any case be subject to the FAT – those mentioned above.
2.5.
Taxable base: the different FAT and methods for calculation of relevant profits
As mentioned, the FAT would be levied on the sum of profit and remuneration for labour of
financial institutions. Technically, there are different ways of defining profit and remuneration
for FAT purposes. The choice of the method depends on the ultimate objective sought with
the introduction of the FAT, namely, taxing value added, taxing economic rents, improving
market efficiency and/or discouraging risk-taking activities.
The figure below illustrates how different methods to calculate the taxable base for FAT
purposes can be conceived to achieve different FAT, which would in turn pursue different
economic objectives.
Figure (2): Variants for a FAT
7
Intermediation fees earned by financial institutions can be considered as business income which means that
they would normally be exempt from any withholding tax at source pursuant to DTCs. Furthermore, in order to
ease the raising of capital by their domestic borrowers, EU Member States have progressively eliminated the
withholding taxes on interest, or certain types of interest.
8
The classification of the taxpayers depending on their turnover is for instance very often used to identify SMEs
(also for direct taxation purposes).
EN
22
EN
FAT
Value added/ economic rent/ return to risk
Profit
CashflowDefinition
2.5.1.
Remuneration
Accrual
definition and
Allowance for
Corporate
Equity (riskadjusted return
for FAT3)
Total
Remuneration
(FAT1)
Remuneration
with deduction
for normal
(FAT2) or riskadjusted return
(FAT3)
Option 2A: The addition method FAT (FAT1)
The addition-method FAT intends to tax value added of financial institutions. Given the
difficulties to determine the value added of financial activities on a transaction-per-transaction
basis, the FAT would tax the aggregated value added of all the financial activities undertaken
by a given financial institution during a certain period of time.
In order to better achieve this, the profit and remuneration of the financial institutions for
FAT1 purposes could be calculated on a cash-flow basis. This cash-flow tax would be
effectively levied on economic rents – this is, from an economic point of view, on projects
with a positive net present value.
Box (8). Cash-flow taxes
Cash-flow taxes have been the subject of a relatively rich amount of economic literature, starting with the
recommendations of the Meade Committee (1978) in the UK (which work is currently being followed-up by the
Mirrlees Review).9 In its simplest definition, a cash-flow tax at the corporate level would tax corporations on the
difference between the sales of goods and services and the purchases of goods and services from other
businesses and from employees. This makes two major differences compared to a classical corporate income tax:
(i) assets are immediately expensed rather than capitalized and depreciated – or re-evaluation and appreciation of
the assets would not be possible either; (ii) sales and purchases are accounted for on a cash-basis and not on an
accrual basis (Bradford, 1986).
Cash-flow taxes are typically assessed on the profit resulting from real – in the sense of non-financial –
transactions (R). However, the nature of the activities of the financial sector makes it impossible to distinguish
the “real” from the financial transactions (F).
9
See e.g. Bradford (2000, 2004), Grubert and Newlon (1997), Bond and Devereux (1995, 2003), Auerbach and
Devereux (2010), or OECD (2007)
EN
23
EN
A cash-flow FAT would be applied on the so-called R+F base.10 This would be formed by (i)
the sum of cash-in from sales, borrowed funds, interest received and loan repayments, minus
(ii) the sum of cash-out from purchases and investments, interest paid, debt repaid and lent
funds. Cash in and out from trading with derivatives or any other financial instruments would
also be considered. In principle, remuneration for labour is not included as a tax-deductible
expense and would consequently be taxed.
Arguably, cash-accounting would lead to lesser differences across EU Member States than
accrual accounting. From a practical point of view, there would be two methods to calculate
the FAT taxable base in cash-flow terms, considering the available data for accrual
accounting:

Under an indirect method, the FAT taxable base in cash-flow terms could be
calculated by adjusting the Profit and Loss Account (P&L) result calculated in
accrual accounting (i.e. the profit as used for the calculation of CIT). In essence, such
calculation would require deducting from the P&L result the full cost of the
investment undertaken by the taxpayer during the year and adding back the
depreciation and the remuneration for labour. Remuneration would be broadly
defined, it would include wages and salaries, bonuses and other performance-related
pay schemes, as well as non-financial advantages (e.g. company cars, company
phones, etc). Please refer to section 6.2.2 and to Annex 11 on revenue estimations to
find an approximation of the indirect method, which has been used for calculating
the revenue potential of the addition method FAT.

Under a direct method, the profit in cash-flow terms would be calculated as the
difference between (i) cash-in from sales and trading of financial products, services
and fixed assets plus the increase in borrowing plus interest income minus (ii) cashout from the purchase and trading of financial products, operating expenses, and
investment in fixed assets, plus the repayments of borrowing and interest paid.
Wages would not be considered as deductible.
2.5.2.
Option 2B: The rent-taxing FAT (FAT2)
The FAT can also be designed to tax economic rents only. Such a tax would be designed by
taxing profit and remuneration for labour above a defined level. The rent-taxing FAT aims at
taxing the rents accruing to the financial sector while leaving untaxed the normal return to
10
A cash-flow corporate tax base can be of 3 different types: the R-base, the R+F base, or the S-base. The Rbase corporate cash-flow includes real transactions only. That is the difference between sales and purchases,
excluding financial transactions (both in terms of revenues or expenses). The R+F base includes real transactions
and non-equity financial transactions (i.e. borrowing and lending of funds). The S-base includes the net flow
from corporations to shareholders. This is dividends paid plus the purchase of shares, minus the issue of new
shares. See OECD (2007). Note that a positive (negative) net result implies a cash-flow to (from) stockholders.
There is therefore an equivalence between the ‘R+F’ base and the ‘S’ base, which consists of net payments to
stockholders (dividends or sales and purchases of shares) (see Bradford 1986, p.120 and following).
EN
24
EN
capital and labour factors. The taxable base of FAT2 could be calculated in the following
ways:

For the profit part, the normal return to capital can be exempted by using the result of
the P&L, adjusted with an Allowance for Corporate Equity (ACE). The Allowance
for Corporate Equity or ACE would allow a deduction for a notional return to equity,
which would be calculated by reference to the interest rate payable on low-risk debt
(government bonds, for instance). Some countries use this method in order to allow a
tax deduction that is designed to be roughly similar to the cost of debt-financing. A
possibility would be to couple this ACE with a restriction of the deductibility of the
interest expense on debt to the same rate, so that an identical normal return on debt
and equity is untaxed. Another method could be one general allowance rate for total
corporate capital (see Annex 18 for more details).

For the wage part, the normal return to labour can be exempted by providing an
allowance for ‘normal’ remuneration. Such ‘normal’ remuneration per employee
could be for instance defined as the average remuneration in other non-financial
sectors, possibly correcting for the type of functions11.
2.5.3.
Option 2C: The risk-taxing FAT (FAT3)
A third version of the FAT would tax excess return due to unduly risky activities. This version
of the FAT is very similar to the rent-taxing FAT because would leave untaxed the 2normal
profit” and the “normal remuneration for labour”. However, the risk-taxing FAT would
determine the 'normal profit' by adjusting the P&L result with a higher ACE (i.e. higher
‘normal’ return for corporate equity), so that only the excessive return to (average) equity is
taxed.
The reason is that this excessive return is arguably the result of high risk-taking activities.
Therefore, parts of the rents could theoretically be untaxed as long as the return to equity does
not exceed this threshold. Similarly to the FAT2, only expenses for remuneration for labour
over a certain threshold would also be added-back to the P&L result and be, consequently,
taxed.
2.6.
The tax rate
The Commission SWD has used a 5% rate for illustrating the revenue collection potential of
the FAT. However, the appropriate FAT rate would depend on the revenue collection
expected and on the other objectives pursued by the introduction of the FAT (mostly on the
FAT taxable base).
An argument could be made in favour of an FAT rate equal to the VAT rate, so that the FAT
best addresses the VAT exemption of financial services. However, to the extent that financial
11
Theoretically, the same objective could be achieved via changes in the CIT base, e.g. by a limitation of the
deduction of remuneration as costs above a certain threshold.
EN
25
EN
institutions may not recover input VAT, such a rate could excessively burden the sector. 12
Nevertheless, linking the FAT rate to the VAT rate applicable in the Member State would be
an option worth considering in the final design of the FAT but this could be hampered by
inconsistencies in the actual impact of non-recoverable VAT (see in particular Annex 17 for a
discussion on these issues).
Economic literature proposing the cash-flow taxation as a substitute for the classical
Corporate Income Taxation notes that the neutrality of the tax with respect to investment
depends crucially on the tax rate being constant over time.13 In addition, a unique, or at least
very similar, FAT rate applicable across the EU would probably be desirable in order to
minimise distortions and to ease the application of the methods to avoid double taxation (both
issues are further addressed below).
In summary, tax neutrality would call for an FAT rate lower than the VAT rate, which would
be as harmonised and constant as possible for all EU Member States. The revenue estimations
provided under section 6.2.2 are calculated using a 5% tax rate for FAT which corresponds to
the minimum reduced VAT rate in the EU.
2.7.
Double taxation issues
The introduction of an FAT, in whatever of the forms described above, would cause different
double taxation issues, which must be distinguished. First, there is the treatment of losses
under FAT and the risk that the profits of a certain financial institution operating in several
EU Member States be taxed twice or multiple times under the FAT of different Member
States. Secondly, there is the interaction between FAT and other existing taxes, notably the
VAT, which are generally referred to as cascading effects, but also with CIT (even in non-EU
Member States). They are addressed below separately.
2.7.1.
Treatment of losses under FAT
Each type of FAT calls for a different treatment of the losses, since each of them would
pursue a different goal:

Addition method FAT or FAT1: symmetrical treatment of gains and losses, this is,
carry-forward of losses should be allowed.

Rent-taxing FAT or FAT2: the carry-forward of losses should be allowed so that the
tax falls on pure economic rents.

Risk-taking FAT or FAT3: Allowing the carry forward of losses would not help
discouraging highly-risky activities, because the losses in one year would be used to
12
IMF (2010a, 2010b)
It is argued that a revenue-neutral reform would require a higher statutory tax rate than the existing CIT rate.
However, to the extent that the FAT would not be introduced as a substitute for CIT but in addition to the CIT,
this argument is not relevant. See Auerbach, Devereux and Simpson (2010) and OECD (2007) p. 102.
13
EN
26
EN
compensate excessive risk-taking in other years. Carry-forward of losses should
therefore not be permitted under the FAT3.
2.7.2.
FAT Double taxation - relief for losses and for FAT in other countries
To the extent that the FAT is introduced in all EU Member States in a coordinated fashion,
FAT double taxation should not be too complex to avoid at least in a purely EU context.
Indeed, the more coordinated the FAT across EU Member States, the less complex would be
to design an effective method to avoid double taxation.
If the FAT is levied on an entity-per-entity level on a stand-alone basis, a method of
imputation or exemption of dividends could be envisaged to relief double taxation in a profitmaking context. This, however, would leave unsolved the taxation in loss-making situations
(provided that a FAT allowing the compensation of losses is adopted). Technical rules on the
treatment of losses made by group entities in different Member States would probably be
needed so that the FAT is as neutral as possible within the EU.
Another possibility would be to have the FAT applied on a consolidated base at EU level. The
current proposal on a Common Consolidated Corporate Tax Base (CCCTB) could serve as a
model to arrive to an EU-harmonised FAT tax base.
2.7.3.
FAT-VAT interaction: cascading effects
The interaction (or rather, the lack of it) between the FAT and the VAT would be similar to
the interaction between VAT and any other direct tax. This is, to the extent that the FAT
would not be assessed on a transaction-per-transaction basis, it would not be possible to have
the FAT properly invoiced on transactions between financial institutions and business subject
to VAT. The financial institutions subject to FAT would not be able to credit input VAT
against their payable FAT (the problem of non-recoverable VAT is further analysed in Annex
17). In addition, business subject to VAT would not have the possibility to credit the FAT for
VAT purposes or to obtain credit for the unrecoverable VAT borne by the financial
institutions from which they have received financial services.
Some technical measures have been suggested to try to alleviate the cumulative effect of
having two different taxes, VAT and FAT, on the same concept (value added). It is has been
proposed to allocate credits for FAT to business on the basis of an approximation, or to have
an FAT rate lower than the VAT rate. None of these solutions seems entirely satisfactory but
further technical work may be carried out to minimise this distortion. As more fully explained
in Annex 17, the potential integration of FAT and VAT would also very much depend on
whether the VAT exemption of the financial services is kept in the current terms or whether a
form of taxing them for VAT purposes is introduced.
EN
27
EN